[
https://issues.apache.org/jira/browse/NIFI-13803?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17885931#comment-17885931
]
ASF subversion and git services commented on NIFI-13803:
--------------------------------------------------------
Commit dffa88aaae6c178a11a888f6cd44a58687243577 in nifi's branch
refs/heads/main from Bryan Bende
[ https://gitbox.apache.org/repos/asf?p=nifi.git;h=dffa88aaae ]
NIFI-13803 Improved Clustered Flow Loading and Synchronization Process
- Only synchronize flow with the coordinator after getting cluster response
This closes #9317
Signed-off-by: David Handermann <[email protected]>
> Setting autoResumeState=false is not honored on a multi-node cluster
> --------------------------------------------------------------------
>
> Key: NIFI-13803
> URL: https://issues.apache.org/jira/browse/NIFI-13803
> Project: Apache NiFi
> Issue Type: Bug
> Affects Versions: 2.0.0-M4
> Reporter: Bryan Bende
> Assignee: Bryan Bende
> Priority: Major
> Fix For: 2.0.0-M5
>
> Time Spent: 10m
> Remaining Estimate: 0h
>
> In a multi-node cluster if you have running components and restart them all
> with autoResumeState=false, the components will still be running on restart.
> This appears to be an issue with how the FlowController is initialized when
> connecting to the cluster. When receiving the cluster's flow and syncing it
> into the FlowController, it uses a scheduler that should be deferring
> scheduling based on whether FlowController is initialized or not. It appears
> that somehow FlowController is considered initialized at this point and when
> synchronizing the flow, any scheduled state changes are taking effect
> immediately, regardless of autoResumeState.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
